The PHAST 2.0 Force Field for General Small Molecule and Materials Simulations

Area of Science:

• Computational Chemistry
• Molecular Dynamics
• Materials Science

Background:

• Classical molecular simulation is crucial for atomic-scale modeling in chemistry.
• Current force fields face challenges in accuracy, efficiency, and transferability.
• Lennard-Jones potentials, common in force fields, have limitations in physical grounding and generality.

Purpose of the Study:

• Introduce PHAST 2.0, a novel general-purpose molecular simulation force field.
• Address limitations of existing Lennard-Jones based force fields.
• Develop a force field with enhanced accuracy, speed, and transferability.

Main Methods:

• Developed PHAST 2.0 with parameters fitted solely to electronic structure data.
• Incorporated explicit many-body polarization and modular many-body dispersion models.
• Explored various atom typing schemes and an implicit polarization version.

Main Results:

• PHAST 2.0 demonstrates accuracy comparable to leading general-purpose force fields.
• Achieved state point independence and reduced complexity with minimal atom typing.
• Showcased emergent generality and transferability, aiding novel chemistry applications.

Conclusions:

• PHAST 2.0 offers a systematic approach to force field development, improving upon Lennard-Jones limitations.
• The method facilitates the creation of bespoke force fields for specific chemical systems.
• This approach enhances computational efficiency and physical grounding in molecular simulations.